Equivalence concept matching is one of major roadblocks towards successful Knowledge Organization System integration. This paper studies equivalence concept matching in the case of thesaurus which is a typical Knowledge Organization System. The related problems and the corresponding solutions are discussed on both effectiveness and efficiency: on the aspect of effectiveness, the reasons for matching errors and matching missing are analyzed, which includes the different preferred terms, term morphology, non-synonyms and homonymies; on the aspect of efficiency, the improvement measures are proposed for the generation of candidate concept matches and manual review. This solution can also be applied to other types of Knowledge Organization Systems. Empirical experiments have been conducted to evaluate the proposed solution on EI thesaurus and INSPEC thesaurus.